Instructions to use MedeHealth/medgemma-chest-xray-v2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use MedeHealth/medgemma-chest-xray-v2 with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("image-text-to-text", model="MedeHealth/medgemma-chest-xray-v2")

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("MedeHealth/medgemma-chest-xray-v2", dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use MedeHealth/medgemma-chest-xray-v2 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "MedeHealth/medgemma-chest-xray-v2"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "MedeHealth/medgemma-chest-xray-v2",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/MedeHealth/medgemma-chest-xray-v2

SGLang

How to use MedeHealth/medgemma-chest-xray-v2 with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "MedeHealth/medgemma-chest-xray-v2" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "MedeHealth/medgemma-chest-xray-v2",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "MedeHealth/medgemma-chest-xray-v2" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "MedeHealth/medgemma-chest-xray-v2",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use MedeHealth/medgemma-chest-xray-v2 with Docker Model Runner:
```
docker model run hf.co/MedeHealth/medgemma-chest-xray-v2
```

MedGemma Chest X-Ray Analysis Model V2

Fine-tuned google/medgemma-4b-it for Portuguese chest X-ray reports.

Usage

from transformers import AutoProcessor, AutoModelForImageTextToText
from PIL import Image
import torch

model = AutoModelForImageTextToText.from_pretrained("MedeHealth/medgemma-chest-xray-v2", torch_dtype=torch.bfloat16, device_map="auto")
processor = AutoProcessor.from_pretrained("MedeHealth/medgemma-chest-xray-v2")

image = Image.open("chest_xray.png").convert("RGB")
messages = [{"role": "user", "content": [{"type": "image"}, {"type": "text", "text": "Analise esta radiografia de tórax."}]}]

text = processor.apply_chat_template(messages, add_generation_prompt=True, tokenize=False)
inputs = processor(text=text, images=[image], return_tensors="pt").to(model.device)

with torch.inference_mode():
    outputs = model.generate(**inputs, max_new_tokens=512, do_sample=False)

print(processor.decode(outputs[0], skip_special_tokens=True))